How the Dexscreener Trending Algorithm Ranks Tokens: Volume, Txns and Unique Wallets

There is a lot of confident nonsense written about the Dexscreener algorithm. People quote exact weightings, invent formulas, and claim insider knowledge. None of that is real. Dexscreener has never published how the board is calculated.

What we do have is behaviour. When you run campaigns across seven chains every day, patterns emerge that hold up consistently. This is what those patterns show.

The three signals that move the board

Every observation points at the same three inputs.

Volume

USD value traded through the pair over the measurement window. This is the loudest signal and the one everyone focuses on, which is precisely why it is not sufficient on its own. Volume is the easiest thing in crypto to manufacture.

Transaction count

The number of individual swaps hitting the pair. This is what separates one whale buying $200,000 from four hundred traders buying $500 each. The same volume, very different signals, and the board treats them differently.

Unique wallets

The number of distinct addresses that traded the pair. This is the signal that matters most for anyone trying to build a position deliberately, because it is the hardest to fake convincingly and the one that separates genuine interest from a loop.

The rolling window, and why the board lags

The board reads activity over a rolling period rather than instantaneously. That single detail explains most of what confuses people about trending.

It explains the delay at the start. You open a campaign and nothing happens for half an hour, because the window has not accumulated enough of the new activity yet to change the ranking.

It also explains why a pair does not appear the instant a window opens. The measurement period needs to fill with the new activity before the ranking reflects it, which takes under an hour on the fast chains and closer to two on mainnet.

Worth separating clearly: a package buys presence on the list for a fixed number of hours. Position within the list is a different thing, decided continuously by what every other pair is doing. How long a placement lasts covers both the bought version and the organic one.

The board is relative, not absolute

This is the part that makes fixed volume promises meaningless.

There is no threshold at which a pair trends. There is only a ranking against every other pair on that chain at that moment. When the whole market is active, everyone's numbers rise and the bar rises with them. On a flat Tuesday afternoon, the same campaign places higher for the same money.

So when a provider quotes you "$X of volume guarantees a top 10 spot", they are either quoting an average and hoping, or they are not thinking about it at all. The right answer to "how much volume do I need" always starts with "on which chain, and on what kind of day". We work through the method in how much volume you actually need.

What does not appear to move the board

Several things people spend money on have no observable effect on trending position:

• Holder count. Useful for trust, irrelevant to the ranking.
• Market cap. A large cap does not place a pair on the board. Plenty of large tokens never trend.
• Social following. Dexscreener does not read your Twitter.
• Boosts. Boosts are a separate paid product that adds a badge and a score. They are not the trending ranking, though they are often bought alongside it. The difference is worth understanding before you spend on either.
• Enhanced Token Info. It makes your pair look complete, which affects conversion, not rank.

That last distinction matters. Several of these things improve what happens after someone finds your pair. None of them help them find it.

Patterns the board appears to discount

Not everything that produces volume produces ranking. From what we observe across campaigns, a few patterns consistently underperform relative to the raw numbers they generate:

• Perfectly regular timing. Swaps at exact intervals are trivially identifiable and do not carry the weight the volume suggests.
• Identical trade sizes. Real trading is messy. Uniform sizes are not.
• Tiny wallet sets. Volume concentrated in a handful of addresses behaves as though it is discounted heavily.
• Closed loops. The same wallets buying and selling to each other repeatedly, with no net position change anywhere.

This is why a properly built campaign costs more than running a bot. The randomisation, the wallet funding, the size variance and the pacing are the actual product. Why volume bots alone do not work goes into what those patterns look like on a chart.

Reading the signals on someone else's pair

The fastest way to understand the algorithm is to stop reading about it and go look at the board. Every pair listed there is showing you its own volume, transaction count and wallet count side by side.

Compare them. A pair with heavy volume and a low wallet count is a concentrated position, and it will usually sit lower than its volume suggests. A pair with modest volume spread across many wallets often sits higher than expected. That relationship, visible in a few seconds of looking, is the algorithm telling you what it weights.

We wrote out the full method in reading the trending page like a trader, because the same three comparisons that reveal how the board works are the ones traders will apply to your pair when your campaign puts it there.

What this means for a campaign

If the three signals are volume, transactions and wallets, then a campaign has to build all three at once, in proportions that look like normal market behaviour rather than an operation.

That is the whole job. Volume is the easy part. Transaction count follows from volume if you size trades sensibly. Wallet diversity is where the cost and the craft sit, and it is the reason the price of a campaign is what it is.

When you understand that, the pricing across the market makes more sense. A cheap campaign is almost always cheap because it skipped the third signal. The price comparison shows where each tier sits and what the savings usually cost you.

If you want the practical version rather than the mechanical one, the main trending guide covers cost, timing and chain selection end to end, and our own campaigns are built exactly on the principles above. The current prices per chain are published, so you can work out what building all three signals costs on your network before committing to anything.